Chandrasekaran, Muthu, et al. "Towards feasible instructor intervention in MOOC discussion forums." (2015).

This paper addresses the challenge of facilitating student learning in Massive Open Online Courses (MOOCs) due to the high student-to-instructor ratio, which limits instructors' ability to engage with discussion forums as they do in face-to-face settings. To support instructors, the paper proposes the use of automated systems to guide interventions in discussion forums. By annotating a large corpus of forum content with a typology of pedagogical interventions, the study enables supervised machine learning models to identify effective interventions. These models could eventually lead to the development of dashboards that prompt instructors on when and how to intervene, and potentially automate these interventions, thereby scaling good pedagogical practices in MOOCs.